Over the past decade, measures of cognitive performance in infancy have been shown to be correlated with cognitive status in childhood and adolescence. Although these measures of attention and recognition memory predict later cognitive status over and above that provided by more traditional sensorimotor infant tests, the absolute level of prediction remains modest. The present application proposes to study the predictive validity of look duration during an infant-controlled visual habituation paradigms, and of novelty preferences assessed after habituation. The investigators propose to assess these variables repeatedly from 3 to 9 months. The inclusion of repeated assessments of look duration will allow for the test of hypothesis that data on the course of developmental change in look duration across that period will (a) increase the predictive validity of the measure over similar measures that have been taken only once or twice during infancy, and (b) provide insight as to the independence of underlying mechanisms or processes that may contribute to this prediction. In addition, heart rate will be measured concurrently with infants' looking during these sessions in order to (a) provide independent evaluation of infants' active engagement/attention in the paradigm, and (b) determine the predictive validity of different phases of infant attention that have been shown to occur during infants looking to visual stimuli. Finally, at each infant assessment, novelty preference will be assessed in a paired-comparison test following habituation. Infant novelty preferences have been shown to be predictive of cognitive outcome in a number of studies over the last decade, but it is not clear whether such prediction is carried by the infant's facility in stimulus encoding, or by the recognition memory components that contribute to the response. The assessment of novelty preference after all subjects have been visually habituated to stimuli theoretically equates for individual differences in encoding, and thus allows us to examine the predictive validity of novelty preferences under conditions where encoding processes are dissociated from memory processes.